Guitar output impedance?

[Ruptor]

It seems to me that choosing the input impedance of a DIY box is a bit of a minefield. I measured the resistance of my Epiphone SG dual Humbucker pick up guitar and it goes from 3.7 K to 470 K depending on the switch, tone and volume control settings. Hoping for some inspiration I looked at the input circuit of my early 70's Orange Matamp OR100 but notice there was a modification on the site that said later models had 220 K resistors added. This mod on the later unit changed the input resistance from 1.068 M down to 182 K  so no enlightenment from that.
I can see that you want a high input impedance like 1 M ideally but in the real world there can be a significant distance from the guitar to the pedal to allow movement on the stage like 10 or 20 foot cable so I wonder if Orange added the resistor to combat noise pick up. Over 200 K would suit my needs and simplify the circuit I want to build since it wouldn't require an extra input buffer and is better than existing Orange Amps being sold. Any thoughts on whether > 200K would be too low or not?

[R.G.]

It seems to me that choosing the input impedance of a DIY box is a bit of a minefield.

Yes. And that is a bit of an understatement.   

I measured the resistance of my Epiphone SG dual Humbucker pick up guitar and it goes from 3.7 K to 470 K depending on the switch, tone and volume control settings.

That's because you're not measuring the pickup itself, but one or more pickups in various combinations and possibly splits, depending on what the switching does, as well as the tone and/or volume controls. And because resistance is not what you need to measure.

Humbucker pickups by themselves mostly measure 4K and up, not far from your 3.7K minimum. That goes through the switches, then the volume pot. When volume is max, the volume pot is in parallel with the pickup, and so you have the volume pot end-to-end resistance (250K to 2M, generally) parallel to the 4K-ish bucker.

But what you need to know is not DC resistance, but what the AC impedance is over the usable frequency range, up to 5K to 10K Hz. Those pickups are actually high-value inductors, usually 2H to 4H for 'buckers, and so their impedance rises linearly with frequency, and is in the range of  100K to 400K at the high end of audio. Amps and pedals will subtract treble more than bass if they're less than about 10x the AC impedance at treble, so you really need on the order of 1M and up in whatever the pickup is driving. In fact, one old trick to fake a pickup is to use a biggish inductor, or a transformer primary of a few Henries to get the same kind of rolloff.

Any thoughts on whether > 200K would be too low or not?

200K is too low unless you like treble rolloff.

[Rob Strand]

You don't have to match the impedance. 
The load affects the high frequency response.

Maybe look here, figs 14 and 15
http://buildyourguitar.com/resources/lemme/
(As written "external" load means external to the pickup not external to the guitar."

At the end of the day use 470k to 1MEG.

[PRR]

Adding resistor will not "combat noise pick up".

Put 220K across your guitar and play. Is the tone OK? Top-end all there? ALL your guitars? Now, and the one you buy next month? Then go for it. Quite a few amps are in this zone. Maybe to cut some harshness (but different cut for different guitars).

The use of 20 30 40 foot cables clobbers highs more than a few hundred K of loading. But guitar-amplifier interfacing has always been a "good enough" bodge, and a 50 foot cable is so cool.

[Ruptor]

Thanks for your replies and thanks to Rob for the link about pickups where the guy has an interesting method of investigating the frequency response of pickups using an external magnetic driver. I looked at some pedals and some are as low as 80K on the input then I tried putting resistors across the guitar lead and watched on the oscilloscope. 47 K knock the signal right down to roughly 25% and 820 K took it to 80% but it says nothing about the frequency content other than I can't see any change in wave shape. Going by what people say when they modify Orange amps from 180 K input load back to 1 M like the original 70's amps the amplifier gain is the only change and there is no mention of any change in frequency content perhaps because it is so insignificant. I guess with a distortion pedal it doesn't matter what it does to the signal so long as it sounds like what you want.

[anotherjim]

When folk speak of low amplifier input impedance not having a noticeable cut of their highs, you have to ask what they are playing with.

They may always have a buffered pedal between guitar and amp. That stops the amp input impedance load for changing the guitar tone.

They may have active pickups or tone control. Ditto as above.

They may have low impedance pickups on a passive guitar. These can be tolerant of load down to 100k or less. Without doing any measurement, the value of volume control in the guitar is a guide. If it's 500k to 1M, then the amp impedance must be as high as is practical - usually 1M. If its 100k to 250k, then it probably can tolerate less that 1M input impedance.

There may be treble loss, but the players amp & cab design or eq choices won't show it up.

The player has severe hearing damage and wouldn't notice treble loss if it walked up and smacked them around the head.

[Ruptor]

They may always have a buffered pedal between guitar and amp. That stops the amp input impedance load for changing the guitar tone. They may have active pickups or tone control. Ditto as above.

Are you saying Orange assume that everybody will be using active guitars or pedals allowing them to drop their amplifier input resistance?

[anotherjim]

More likely someone tested it with a Telecaster or Strat & didn't think to try a Les Paul.

[Rob Strand]

I looked at some pedals and some are as low as 80K on the input then I tried putting resistors across the guitar lead and watched on the oscilloscope. 47 K knock the signal right down to roughly 25% and 820 K took it to 80% but it says nothing about the frequency content other than I can't see any change in wave shape.

Normally, I wouldn't expect that much attenuation, maybe a few dB.  You might be seeing changes in peaks as opposed to the perceived level.   The 470k DC resistance you measured before might also have something to do with it.

Going by what people say when they modify Orange amps from 180 K input load back to 1 M like the original 70's amps the amplifier gain is the only change and there is no mention of any change in frequency content perhaps because it is so insignificant.

Normally it would just give you a little more highs.   For most pickups 180k would have an audible roll-off.  At about 330k it's hard to tell. And 1MEG is usually a good upper bound.   If you consider the guitar has a 250k volume control and a 250k tone control in parallel with it, the pickup is seeing at least 125k, loads like 1MEG only change the total load by a small amount, like 10% - that's why you can't hear much change beyond 1MEG.

I guess with a distortion pedal it doesn't matter what it does to the signal so long as it sounds like what you want.

For effects like an echo, phaser or a chorus, you want the clean signal to come through so you don't want to load the pickup.  But a distortion pedal you can do what you want.

[PRR]

> resistors across the guitar lead and watched on the oscilloscope

Your ears/brain are much smarter than a 'scope!

Bass is big. Treble is small. You can cut everything over 800Hz and hardly see a difference in the total. But your ears will feel like they are full of wool. Very dull sound. No sing no zing.

*Listen* with various loadings. As you come down from a Meg, the treble goes out first. Playing through a blanket. Because much guitar technique is about managing those higher partials (and because you need the scream to cut through a band), you really want to preserve these highs as much as possible.

[thermionix]

The 470k DC resistance you measured before might also have something to do with it.

Yeah you shouldn't see a resistance anywhere near that across the output of any common, passive pickup guitar.  Usually below 20k.  You might have an open pickup, and you're reading a volume pot with nothing in parallel.  It can still making sound, thin and weak, via a capacitive coupling in the coil.

If you have active electronics though, anything's possible.

[anotherjim]

My first thought, for a reading of 470k from the jack socket, is that you have a guitar with one pickup selected by a switch (or only one pickup fitted) and a 500k volume pot turned most of the way down.

[thermionix]

Volume pot all the way down would be "zero" ohms across the output jack.

500k volume pot all the way up is 500k ohms in parallel with the dc resistance of the pickup, usually 6-16k or so.  In a common gibson type guitar, one pickup on.

[antonis]

500k volume pot all the way up is 500k ohms in parallel with the dc resistance of the pickup, usually 6-16k or so.

But what actually minds is the AC resistance of the picup(s)..

[anotherjim]

Volume pot all the way down would be "zero" ohms across the output jack.

No, not if its wired pickup to wiper of volume pot, which some guitars are - mostly 2 pickup/separate control guitars but some single pickup versions do the same, even though they could be wired wiper to jack. Think about it, if 2 volume pots are wired wiper to jack, then turning one down would turn the other down too. Although there is probably more treble loss with low volume control settings (due to the lower wiper resistance to ground loading the treble out), some manufacturers and players prefer the lack of interaction between controls by wiring it that way. With 2 pickup bass guitars, pick up to volume wiper is pretty much the standard way - Fender, Rickenbacker etc.



Now, the OP said it's an Epiphone humbucker SG, which almost certainly means 500k pots. The Epi standard (according to web searches) appears to be volume wiper towards the jack (via the selector). To read 470k at the jack would require no pickup at all selected, which isn't likely unless modified with additional switching that can do that. Either that 470k is an "anomalous" reading, or its wired pickup to wiper and the volume is turned down.

[thermionix]

Well, let me re-phrase.  In any factory-wired Gibson or Epiphone "standard" LP/SG/similar that I'm aware of, a 470k DC resistance reading across the output jack would indicate an open or disconnected pickup.

Wiring the pickup hot to the volume pot wiper is a noise problem, and fairly rarely done.  But it's certainly possible.

[Rob Strand]

it goes from 3.7 K to 470 K depending on the switch, tone and volume control settings

In any factory-wired Gibson or Epiphone "standard" LP/SG/similar that I'm aware of, a 470k DC resistance reading across the output jack would indicate an open or disconnected pickup.

Maybe we need confirmation from the OP.  From what I can see that's the case.  However, across all Gibson guitars both connections are used.

Wiring the pickup hot to the volume pot wiper is a noise problem, and fairly rarely done.  But it's certainly possible.

It's pretty common for basses.   Noise isn't an issue because as you tend to have one of the pickups cranked to full.  You can also back one pickup-up off entirely without killing the signal (and there's no need for a pu selector.)

With a 3-way switch, in the middle position, the wiper to wiper connection causes the two volumes controls to affect each other.

[anotherjim]

My 70's Gibson SG was wired pickups to volume pot wipers. I changed it, didn't like the interaction thing and put it back.

Rickenbacker 4001/3 wire pup to wiper and have a 3 way selector.

Yes, doing it that way does mean that noise can increase when turning volumes down, since the cable to the amp gets a much higher source impedance and worse with 500k pots than 250k's. This point may be very pertinent to the OP if his pickups are wired to the wipers.

[Ruptor]

I don't know how the pickups are wired but they are not open circuit since the switch selects either one or both and you can hear the difference in sound from each of the pickups. The limit values are the lowest and highest resistance values I could measure by turning the pots to the extremes so I guess when the Pots are at maximum the one pickup is directly across the output since you say the pickups are 500K.

[thermionix]

I guess your pickup hots are wired to the volume wipers.  I would not expect that from a factory-wired Epiphone SG.  Anything is possible, and it may have been changed by someone.  I would recommend moving p/u hots to lug 3 (and output to wiper) so you have less hum and noise when you roll your volume off.